NOTES
492 30
i
Vol. 7 7
Piperidine Derivatives. MVI. 1-Methyl-3-benzal-4-piperidone Dimer BY S. M. MCELVAIN AND P. HAROLD PARKER, JR.* RECEIVED AUGUST18, 1954
The dimeric structure I is now assigned to the compound described as l-methyl-3-benzal-4piperidone (11) in paper XVIII of this series.* This compound was obtained in the earlier work from 1-methyl-4-piperidone and benzaldehyde as well as from l-methyl-3,5-dibenza1-4-piperidone (111) when the reaction was carried out in a solvent (470 solution of potassium hydroxide in 60% ethanol) from which 111 did not precipitate. The structure I1 was assigned on the basis of the elemental analyses and the methods by which it had been prepared, although some concern was expressed over the much lower solubility of the monoI I benzal derivative I1 as compared to that of the 20 40 60 80 100 dibenzal derivative 111; also the melting point of Minutes. the former compound (224') is considerably higher Fig. 1.-Oxygen consumption by coniferyl alcohol in than that of the latter (117°).3 the presence of mushroom extracts. Each Warburg vessel In the present work doubt was cast on the struccontained 4 +moles of coniferyl alcohol dissolved in 2.8 ml. ture I1 when it was found that (a) the compound of 0.1 116 phosphate buffer, PH 6.8, in addition to, (A) 4 could not be converted to the dibenzal derivative mg. of Freudenberg oxidase (160 catecholase units), (B) 111, although the reverse transformation is possible2 4 mg. of Freudenberg oxidase heated to 100' for 3 minutes, and (b) the hydrogenation of the compound did not or 160 catecholase units of purified mushroom polyphenol- yield the expected l-methyl-3-benzyl-4-piperidone, oxidase, or 3 pmoles of cupric ion. The dotted line indicates but a compound with a wide melting range (85an amount of oxygen equivalent to 1 / 1 atom per molecule 95') that showed a molecular weight of ca. 390. of coniferyl alcohol. KOH was present in the central well. The structure I accounts for these results as well as for the following observations : the compound tion in accord with the earlier work. However, in (a) shows a molecular weight of 408 (calcd. for I, the presence of the same number of catecholase 402), (b) has a broad associated hydroxyl band units14 of purified mushroom polyphenol~xidase'~joining the C-H band a t 3.3 p and also a hydrogen the oxygen consumption was negligible, as was the bonded, conjugated carbonyl band a t 6.1 p in the case in the presence of boiled Freudenberg oxidase, infrared s p e ~ t r u m (c) , ~ shows a maximum absorpor cupric ion. tion a t 290 m p (log e 4.43) in the ultraviolet specWe conclude that there is present in crude mush- truma and (d) yields an acetate in whose infrared room extracts a heat-labile system which catalyzes spectrum there is no hydroxy band and in which the consumption of some oxygen by coniferyl alco- the a,&unsaturated carbonyl band, which is hol, but it is not polyphenoloxidase which does this. apparent in the ultraviolet spectrum, is masked by The present experiments do not rule out a biosyn- the ester carbonyl band a t 5.75 p . thetic relationship between the lignins and melanins 0---H-0 but they do show that the particular process proII I posed by Freudenberg is not pertinent to the concept.
"..I :
Experimental Coniferyl Alcohol.-The coniferyl alcohol was prepared by the lithium aluminum hydride reduction of ethyl acetoferulate using a modification of the method of -411en and Byers.16 The ester (0.025 mole) was added in one portion to a solution of lithium aluminum hydride (0.100 mole) in The temperature was allowed 800 ml. of ether a t -60". to rise t o 0' over a period of 1.5 hours and the solution was then filtered and the salt was hydrolysed and crystallized as described.16 This method gave a product which crystallized immediately and without difficulty from the petroleum ether-ether mixture. DEPARTMENT OF BIOCHEMISTRY OR OREGON MEDICAL SCHOOL UNIVERSITY A N D THE DEPARTMENT OF CHEMISTRY REEDCOLLEGE PORTLAND, OREGON (14) W. H. Miller, M. F . Mallette, L. J . Roth and C. R. Dawson, THISJOURNAL, 66, 514 (1944). (15) The Treemond Company, New York,supplied our preparation, which contained 6200 catecholase units per ml., or 1510 units per mg. (16) C. F. H. Allen and J. R. Byers. Jr.. THIS JOURXAI.. 71. 2688 (1949).
1
--
CH*
11
It is poqsible that the high melting point of I and its insolubility in non-polar solvents may be due to a salt-like character resulting from the interaction of the carbonyl group which is posi(1) Wisconsin Alumni Research Foundation Research Assistant (1953). 70, 1820 (1948). (2) S. M . McElvain and K Rorig, THISJOURNAL, (3) I t may also be noted that the analog of 11, 2-benzalcyclohexanone, is reported, (D. Vorlander and K Kunze, Bcr., 19, 2078 (1926). to melt at 54O and can be converted quantitatively to 2,6-dibenzalcyclohexanone, m.p. 118'. (4) This compound, a liquid, b p 104-105° (0.1 mm.), has been prepared in this Laboratory by Martin D. Barnett by the decarbethwhich was oxylation of I-methyl-3-benzyl-3-carbethoxy-4-piperidone, obtained by the benzylation of l-methyl-3-carbethoxy-4-piperidone. (5) Cf.R. S. Rasmursen, D . D. Tunnicliff and R R. Brattain THIS JOURNAL, 71, 1068 (1949). (6) Cf.L. Dorfman, Chcm. Rcos.. 65, 47 (19533
Jan. 20, 1955
NOTES
493
examined with respect to its behavior in the Friedel-Crafts reaction and toward aqueous ammonia. Experimental8 If chlorendic anhydride is regarded as a subDimer of l-Methyl-3-benzal-4-piperidone (I).-The following improved method was used for this product. TO stituted succinic anhydride, i t can be expected to 1.13 g. of 1-methyl-Ppiperidone and 1.06 g. of benzaldehyde act as an acylating agent in the Friedel-Crafts in 2 ml. of 95% ethanol was added 0.3 g. of potassium hy- reaction.6 Its non-chlorinated analog, bicyclodroxide in 1 ml. of water. The solution was stirred and [2.2.1]5-heptene-2,3-dicarboxylicanhydride, howheated at 60-70" for four hours. The precipitate was filtered and triturated with water. When dried the light ever, does not react with benzene and aluminum yellow solid weighed 1.47 g. (73%), m.p. 193-207'. Re- chloride, whereas the corresponding saturated ethanol afforded a n analytical compound d0es.l crystallization from 9;'. sample, m.p. 224-225 . The product I showed a strong It has now been established that chlorendic tendency to form supersaturated solutions in ethanol, but recrystallization could be achieved by dissolution in a mini- anhydride reacts with benzene and toluene in the mum of hot ethanol followed by distillation of about two- presence of aluminum chloride to give the expected thirds of the ethanol. This product had the same analyses keto acids in yields of 8549%. tively polarized by hydrogen bonding with the basic nitrogen of another ring.l
and its melting point was not depressed by a sample pre0 0 pared in the earlier work;s mol. wt. (Rast) 408. c1 II c1 /I Hydrogenation of 1.-To 2.01 g. of I was added 11.9 ml. of 0.839 N hydrochloric acid and 0.15 g. of Adams platinum oxide catalyst. The solution was shaken with hydrogen a t 38 pounds pressure until two equivalents of hydrogen per mole of I had been absorbed (65 minutes). The solution was filtered to remove the catalyst and neutralized with so0 dium hydroxide t o give a white solid with wide melting I 11 range (85-95'), which could not be narrowed by recrystallization; mol. wt. (Rast) 390. The acids I1 (R = H, Me) can be esterified but Acetate of 1.-To a mixture of 2.01 g. of I and 0.75 g. of fail to cyclize on heating with concentrated or fused sodium acetate under nitrogen were added 10 ml. of acetic anhydride. The mixture was heated at 110' with 'fuming sulfuric acid, which was also observed in stirring for two hours, cooled and poured into 75 ml. of ice- the case of the analogous 3-benzyl-norcamphanewater. After standing for one hour the aqueous solution was neutralized with sodium bicarbonate. The resulting 2-carboxylic acid.' The crystalline imide of chlorendic acid may be fluffy, white precipitate was washed with water, dried, dissolved in ether and filtered. The ether solution was poured prepared in excellent yield by the reaction of onto a 38-cm. column of Alcoa activated alumina F-20. chlorendic anhydride with aqueous ammonia a t Elution of ether-isopropyl alcohol solution (1004.25) gave 1.3 g. of solid which was triturated with isopropyl ether and room temperature. Experimental* recrystallized from ethyl ether t o give an analytical sample of the acetate of I ; m.p. 170-172'. Chlorendic Anhydride.-Commercial chlorendic anhyAnal. Calcd. for CZ&&ZO~: C, 75.64; H, 7.26. drideQ was purified by refluxing with acetyl chloride for 2 hours immediately before use. After cooling the mixture Found: C, 75.38; H , 7.10. This acetate of I had an ultraviolet absorption maximum was filtered with suction, washed twice with anhydrous a t 287 m r (log e 4.43) and molecular weight (Rast) 455. ether and warmed briefly on a steam-bath to remove reFurther elution of the column with ether-isopropyl alcohol sidual ether. 3-Benzoyl-l,4,5,6,7,7-hexachlorobicyclo [2.2.1]5-heptene(10: 1) gave 0.5 g. of an oil that failed to crystallize. 2-carboxylic Acid.-Chlorendic anhydride (37.1 g., 0.1 (7) For examples of interaction between a nitrogen and a carbonyl mole) was dissolved in 150 ml. of dry, thiophene-free bengroup see N. J. Leonard. R . C. Fox, M. Oki and S. Chiavarelli, THIS zene. To this solution was added all a t once 38.5 g. (0.29 JOURNAL, 76, 630 (1954). mole) of pulverized aluminum chloride. An immediate exo(8) Spectra were determined with a Cary Ultraviolet Recording thermic reaction was controlled by cooling. After the reSpectrophotometer, model l l M S and a Baird Double Beam Infrared action had subsided, the mixture was refluxed gently for Recording Spectrometer, model B, fitted with a sodium chloride prism. one hour. After cooling, the mixture was hydrolyzed by Ultraviolet adsorption spectra were determined in ethanol, Infrared cautious addition, with stirring, of 50 ml. of water, followed adsorption spectra were determined in a Nujol mull. by 100 ml. of 10% hydrochloric acid. The mixture was allowed to stand overnight and steam distilled to remove the DEPARTMENT OF CHEMISTRY benzene. The solid, which remained in the residue, was OF WISCONSIN UNIVERSITY filtered with suction and digested with two 400-ml. portions 6, WISCONSIN MADISON of boiling water. The remaining tan solid was decolorized with Norit and crystallized from benzene. Drying of the crude, m.p. 187-192", a t 130" and recrystallizing from the Some Reactions with Chlorendic Anhydride' same solvent gave colorless prisms, m.p. 196-197', yield BY E. T. MCBEEAND W. R. DIVELEY~ 85%. Anal. Calcd. for ClsH80aCls: C, 40.09; H, 1.78; neut. RECEIVED JULY 21, 1954 equiv., 449. Found: C, 40.12; H, 2.07; neut. equiv., 446. Ethyl Ester.-On esterification with ethanol and sulfur: I n connection with other investigations in this Laboratory concerning hexachlorocyclopentadiene acid, the acid was converted to the ethyl ester, m.p. 80-81 Anal. Calcd. for CnH1zClsOa: C, 42.80; H, 2.53. and its derivativesja-5 chlorendic anhydride was Found: C, 42.74; H, 2.21. (1) The simple name "chlorendic anhydride" for 1,4,5,6.7,7-hexa1,4,5,6,7,7-Hexachloro-3-(p-toluyl)-bicyclo [2.2.1] 5-hepchlorobicyclo [2.2.115-heptene-2,3-dicarboxylic anhydride conforms t o tene-2-carboxylic Acid.-The analogous reaction of 37.1 g. the International system of nomenclature and t o the system of Patter(0.1 mole) of chlorendic anhydride, 150 ml. of toluene and
.
son and Capell in the "Ring Index" (Chem. Eng. News, S a , 873 (1954)). (2) Abstracted from the doctoral thesis of W. R . Diveley, Purdue University, 1952. (3) E. T. McBee and C. F. Baranauckas, I n d . Eng. Chcm., 41, 806 (1849). (4) J. S. Newcomer and E. T . McBee. THISJOURNAL, 71, 946, 952 (1949). ( 5 ) E.T . McBee. H. Rakoff and R. K. Meyers, ibid., 77, in press.
(6) R . Adams, "Organic Reactions," Vol. V, John Wiley and Sons, Inc., New York, N. Y., 1949, p. 229. (7) M. S. Morgan, R . S. Tipson, A. Lowy and W. E. Baldwin, THIS JOURNAL,66, 404 (1944). (8) All temperatures are uncorrected. (9) The authors are indebted t o the Hooker Electrochemical Company for a generous supply of this anhydride.